The original version of this document was dated November 13, 2015.
Version 2: January 7, 2016.
OSFI has received a number of questions regarding the Quantitative Impact Study #7 (QIS#7). Below are the answers to the most commonly asked questions, organized by major risk category.
Corrections or clarifications will be made, if necessary (in blue), to original questions and answers from earlier versions of this document. The dates of the changes or additions will be indicated in brackets. New questions and answers will be appended below under the version posted, if needed.
G.1) For Operational Risk, could you please specify if the exposure for annuities (gross of reinsurance liabilities) is best estimate liabilities or padded liabilities?
The exposure for annuities should be based on padded liabilities.
G.2) Summary Comments and Questions Workbook – English. In Diversification Credit worksheet, Insurance Risk PfAD for Lapse (row 18) is not linked to the correct row, i.e. it is currently linked only to the Par PfAD and not the Par + NonPar PfAD.
Please make the following corrections to the Diversification Credit worksheet:
- Cell E18 should link to ActLiab_PfAD_CALMBestEst!$M71;
- Cell F18 should link to ActLiab_PfAD_CALMBestEst!$M71;
- Cell G18 should link to ActLiab_PfAD_CALMBestEst!$M131;
- Cell H18 should link to ActLiab_PfAD_CALMBestEst!$M161;
- Cell I18 should link to ActLiab_PfAD_CALMBestEst!$M191;
- Cell J18 should link to ActLiab_PfAD_CALMBestEst!$M221.
G.3) Summary Comments and Questions Workbook – French. Diversification Credit worksheet has various errors.
A revised Summary Comments and Questions workbook (version 2) with an updated Diversification Credit worksheet has been posted to OSFI’s website.
M.1) Appendix I in the Market Risk instructions needs to be updated for Participating under the Base Scenario where CALM base scenario dividends should be used.
In Appendix I, Market Risk – Interest Rate Risk Cash Flows, the phrase “Dividends use QIS forward rates as investment returns” in the Base Scenario / Participating cell should be replaced with “Dividends are CALM base dividends”.
M.2) At the bottom of page 8 of the Market Risk instructions, the last paragraph reads: “Asset and liability cash flows associated with amounts on deposit, including from participating dividends, should be treated as time zero cash flows”. This poses a practical challenge as assets backing the amounts on deposit are not separated, hence the amount to be held at time zero, and will have to be approximated.
The Deposit Liabilities section of the instructions should be deleted entirely and be replaced with the following: “The balance sheet value of deposit-type liabilities that are classified as investment contracts in the financial statements should be projected as a time zero cash flow”.
The instructions on deposit liabilities should not make any reference to assets. Assets should be projected based on their contractual cash flows. All relevant asset cash flows should be included as investment contract cash flows in the Asset and Liability Cash Flows worksheet of the Market Risk workbook. If it is not possible to disclose the asset cash flows as investment contract cash flows, then these asset cash flows should be included with another category according to your asset-liability matching practice and disclosed in the Questions and Comments worksheet.
Due to the application of the above instructions, an explicit interest rate risk solvency buffer will not be required for deposit liabilities in the Summary worksheet of the Market Risk workbook. Therefore, Market Risk workbook, Summary worksheet, cell M11 through to cell X11 should not contain any values, i.e. should be zero.
M.3) Limited Partnerships. Page 3 of the Market Risk instructions states the following:
"...For the QIS solvency buffer, insurer should use the same approach and methodology as the current MCCSR Guideline (with a factor of +/-35% instead of +/-15%, excluding segregated funds) for the following:
- Limited partnerships..."
However, page 15 of the Market Risk instructions (Equity Risk section) states that "...The look through method does not continue to apply for limited partnerships investments subject to MCCSR section 3.1.12."
Is it the intent that the current capital approach applies to limited partnerships in QIS#7 except for the look through method?
The look through method in section 3.1.12 of the 2014 MCCSR does not apply in QIS#7. Insurers should treat investments in limited partnerships that are not substantial investments (as defined in section 10 of the Insurance Companies Act (ICA)) as equity investments and apply a factor of 35%. A factor of 45% is applied to substantial investments in limited partnership without control (net of the amount of related goodwill). Control is defined in section 3 of the ICA. Controlled investments in limited partnerships should be consolidated and the assets treated according to their nature (e.g. real estate).
Therefore, the instructions on page 3 should not refer to limited partnerships and the instructions defining “equities” in the Equity Risk section on page 15 should be amended to include limited partnerships and read as follows: “all investments classified as equities (including equity index securities, managed equity portfolios, investment trusts in equity and limited partnerships), excluding preferred shares, should be calculated using a simple immediate (time zero) deterministic downward shock of 35% to the market value of the equities.”
M.4) Cash Flows linked to Inflation. Page 5 of the Market Risk instructions states that "...Future maintenance expenses should not be adjusted for inflation". Page 24 of the Insurance Risk instructions states: "The shock for level, trend, volatility and catastrophe risk is an increase of 20% in the first year... on the best estimate expense assumptions including inflation..." Is it the intent that the inflation application on maintenance expenses is different between interest rate risk and insurance risk?
The sentence on page 5 of the Market Risk instructions should read as follows: “The inflation assumption on future maintenance expense should not be adjusted based on scenarios but should be the same as used in the base scenario.” The inflation assumptions used in the interest rate risk and insurance risk should be the same.
I.1) Lapse catastrophe risk.The factor has increased from 20% to 30% for level and trend risk, but for the catastrophe risk, the shock for lapse sensitive product continues to be +20%. Is this intentional?
The increase from 20% to 30% (e.g. for the level and trend risks) is a shock applied to best estimate assumptions (e.g. 5% + (30% x 5%) = 6.5%). The catastrophe risk shock of 20% is applied in addition to the best estimate assumption (e.g. 5% + 20% = 25%) and has not increased.
I.2) Lapse volatility risk.There is a new sentence in the instructions: “The cross over logic should be used (i.e. shock should be independent of the designation)”. It will be a challenge to model cross over logic for a one year shock due to software limitations.
Cross over logic is not used in the designation process of QIS#7. However, when applying the shocks for level and trend, the instructions specify that the insurer should apply the lapse shocks in a manner consistent with how the margins for adverse deviations (MfADs) for lapse are applied for valuation purposes as per the CIA Educational Note on Margins for Adverse Deviations [document 206132], which includes cross over logic, ensuring that the resulting buffer is positive in aggregate for each portfolio. It also addresses immaterial situations and exceptions.
If the volatility buffer cannot be calculated due to software limitations, then the volatility buffer may be approximated by applying the volatility shock by portfolio, while ensuring that the resulting solvency buffer for the portfolio is positive. Approximations used should be disclosed in the Questions and Comments worksheet of the Summary Comments and Questions workbook. Please disclose the materiality of these impacted portfolios (% of the portfolio in comparison to the total liability).
I.3) CALM discount rates used for designation test. "CALM discount rates" is mentioned in a few places in the QIS instructions. For example, the instructions state that CALM discount rates should be used to determine the life/death supported and lapse sensitive/lapse supported designation:
- Mortality Risk – Designation test for life and death supported business (IR instructions p. 8):
"The calculation requires taking the present value using CALM base scenario discount rates of -15% level risk with +75% trend risk. Insurers should compare the results of the calculations to the present value using CALM base scenario discount rates of the CALM best estimate cash flows. If the result of the calculation is greater than the present value of the best estimate cash flows, the business is designated as death supported otherwise the business is designated as life supported."
- Lapse Risk – Designation test for lapse supported and lapse sensitive business (IR instructions p. 21):
"...The designation is done on a portfolio basis based on the largest present value using CALM discount rates, even if the present value under each test is less than the best estimate present value.”
For products without CALM discount rates and any other inferred rates, what discount rates should the designation test be performed with? For example, there are some blocks that currently set the liabilities using an approach other than CALM or PPM (i.e. percentage of account value).
For products without CALM discount rates and any other inferred rates, the designation used in the CALM valuation may be used and disclosed in the Questions and Comments worksheet of the Summary Comments and Questions workbook.
C.1) Calculation of Effective Maturity. In the instructions it mentions that all exposures to a connected group should be aggregated before calculating the effective maturity date. Please clarify if an average effective maturity date should be calculated for all exposures to each connected group?
Aggregation is applied to all obligations due from entities within a connected group within each rating grade. Hence a single maturity is calculated based on the aggregated cash flows of all obligations due from such entities.
C.2) Assets Receivable and Recoverable – Counterparty Risk. In QIS 7 Credit Risk Instructions about the Counterparty Risk for Assets Receivable and Recoverable (page 7), it states:
“The factors for outstanding premiums, including installment premiums receivable, from other than FRI’s and approved provincial reinsurers are:
- 0% for installment premiums receivable (not yet due);
- 5% for premiums outstanding less than 60 days, including installment premiums receivable;
- 10% for premiums outstanding 60 days or more, including installment premiums receivable.”
We are not sure what " installment premiums receivable (not yet due)" refers to. Can OSFI clarify what kind of installment premiums receivable are eligible for the 0% factor category?
The term “Installment premiums receivable (not yet due)" refers to installment premiums not paid after the payment date, but before the end of the grace period. For example, the 0% factor would apply to a policy with an annual or monthly premium payment date of December 15, not paid on December 31, but with a grace period of 30 days.